Expect the randomized ~E tests FORMAT.E.1 and FORMAT.E.2 to fail sometimes

[dg1sbg]

FORMAT.E.1 and FORMAT.E.2 fail intermittently on unmodified main and on unrelated pull requests, and each failure costs someone an adjudication before the red can be dismissed. They are not deterministic ~E tests. Each draws a float type with (RANDOM 4), draws a random exponent, repeats 1000 times, and compares (FORMAT NIL "~E" X) against PRIN1-TO-STRING and against a compiled FORMATTER call (dependencies/ansi-test/printer/format/format-e.lsp:12,45). Whether they trip is a property of the draw.

What they catch when they do trip is the round-trip float-printing defect already tracked as #1816, reached through ~E rather than through PRIN1: the two printers disagree about where the decade boundary falls when the mantissa rounds up to it. On a build of 0aa5b71ac, ~E prints the single-float 9.99999e-18 as "0.999999e-17" while PRIN1 prints "9.99999e-18", and the double-float 9.9999999e-18 disagrees the same way. It is not specific to LONG-FLOAT and not specific to extreme exponents; any of the four float types will do, at ordinary magnitudes.

Recent instances: main at 0aa5b71ac red on FORMAT.E.1 with that exact single-float value, and #1820 red on FORMAT.E.2 with -9.999338424609615l-2430. In both cases it was the only unexpected failure out of 21936, with the regression suite green in the same job.

Why listing them is free

DO-ENTRIES returns (null *unexpected-failures*) (rt.lsp:521), so a listed test that passes is recorded as an unexpected success (rt.lsp:496) and reported, but does not change the exit status. Both names match loaded test entries, so neither adds to the Found unknown test or note names report — verified by loading init.lsp and calling rt::load-expected-failures against the edited file:

UNKNOWN-EXPECTED-FAILURES: NIL
FORMAT.E.1: symbol=T known-test=T listed-as-expected-failure=T
FORMAT.E.2: symbol=T known-test=T listed-as-expected-failure=T
TOTAL-EXPECTED-FAILURES: 32

This is already how PRINT.LONG-FLOAT.RANDOM, PRINT.SYMBOL.RANDOM.3, PRINT.SYMBOL.RANDOM.4 and FORMAT.E.26 are handled; FORMAT.E.26 is defined in the same file and already carries the fails sometimes note.

Scope

This does not fix #1816, and it deliberately leaves PRINT.SHORT-FLOAT.RANDOM, PRINT.SINGLE-FLOAT.RANDOM and PRINT.DOUBLE-FLOAT.RANDOM unlisted even though PRINT.DOUBLE-FLOAT.RANDOM has also been seen red on main. Those are a judgement for whoever owns the printer work; the smallest change that stops the recurring noise is the one worth making now. Happy to widen it if you would rather have the whole randomized-printer family listed at once.

Refs #1816

[dg1sbg]

Widened, and the CI run on the first version is the reason.

Run 32345271440, job 96352593574 — clasp/ubuntu-latest/native/yes/no, step Run ANSI tests:

29 failures with 1 unexpected failure and 4 unexpected successes out of 21936 tests.
Unexpected failures:
  PRINT.DOUBLE-FLOAT.RANDOM
Unexpected successes:
  PRINT.SYMBOL.RANDOM.3
  PRINT.SYMBOL.RANDOM.4
  FORMAT.E.1
  FORMAT.E.2
FAILED: [code=1] ansi-test-boehmprecise

That single job contains both halves of the argument. FORMAT.E.1 and FORMAT.E.2 drew clean, landed in unexpected successes, and contributed nothing to the exit status — the mechanism works exactly as described. And the job still went red, on PRINT.DOUBLE-FLOAT.RANDOM: the test I had deliberately left out and flagged in the description as also known to red on main. The next draw from the same family reddened the next cell.

So this now also lists PRINT.SHORT-FLOAT.RANDOM, PRINT.SINGLE-FLOAT.RANDOM and PRINT.DOUBLE-FLOAT.RANDOM, joining PRINT.LONG-FLOAT.RANDOM, which was already there. Those four are the same test written once per float type — each walks the exponent range from least-positive to most-positive, draws ten values per decade with (RANDOM f) and feeds them to RANDOMLY-CHECK-READABILITY (print-floats.lsp:86,181,276,371, checking at :100,195,290,385). Listing one of the four and not the other three was arbitrary; the underlying round-trip defect (#1816) is not specific to a float type.

Re-verified against the loaded test entries before pushing, same way as before — no unknown names, 32 → 35 expected failures:

UNKNOWN-EXPECTED-FAILURES: NIL
PRINT.SHORT-FLOAT.RANDOM:  known-test=T listed=T
PRINT.SINGLE-FLOAT.RANDOM: known-test=T listed=T
PRINT.DOUBLE-FLOAT.RANDOM: known-test=T listed=T
PRINT.LONG-FLOAT.RANDOM:   known-test=T listed=T
FORMAT.E.1:                known-test=T listed=T
FORMAT.E.2:                known-test=T listed=T
TOTAL-EXPECTED-FAILURES: 35

Still no attempt at #1816 itself — this only stops the randomized printer tests from reddening unrelated PRs. Happy to trim it back if you would rather keep some of these visible.